chat-with-gpt

Why we included this project

This is a self-hosted chat client that puts a more configurable face on the OpenAI API than the standard ChatGPT page. You supply your own API key, and in return you get searchable history, editable and regenerated messages, a temperature slider, full markdown rendering, and shareable URLs for individual conversations. The standout feature is voice: it can read replies aloud through ElevenLabs or the browser's built-in speech engine, and accept spoken input via OpenAI's Whisper, which makes hands-free use genuinely practical. It runs from a single Docker command, stores keys either on your device or your server, and lets developers spin up a private, customizable front-end in minutes. For teams curious how an OpenAI-backed assistant feels as a real product UI rather than a raw API call, it's a solid reference.